Why three measures of the same light don't tell the same story
Melatonin and cortisol are hormones used as biochemical markers of the body's circadian phase: melatonin rises in the evening, and cortisol peaks shortly after waking [2]. EEG, by contrast, records electrical activity from the scalp over seconds to minutes. A fluorescent lamp can plausibly affect all three, but each measure integrates light exposure over a very different timescale, so a change in one need not appear in another at the same moment [1][2].
The hormone side also carries its own measurement burden. Melatonin and cortisol can be quantified in blood, saliva, or urine using immunoassays or mass spectrometry, and results shift with ambient light, body posture, and exact sampling times [2]. That means a hormone difference between two lamp conditions can reflect sampling conditions as much as the lamp itself, which makes it hard to treat hormone shifts and EEG shifts as confirming each other [2].
What the lamp study actually measured
The study closest to the question compared two fluorescent lamp types at two light intensities in healthy humans and recorded melatonin, cortisol, EEG, ECG, and subjective comfort [1]. Its abstract frames this within the broader EEG literature on non-image-forming light responses, meaning light effects that run through pathways other than conscious vision [1]. So the design does put all three measures in the same room under the same lamps, which is exactly the setup needed to ask whether they converge.
What the available summary does not provide is a result showing the three measures agreeing or disagreeing in a specific direction. The abstract-level evidence establishes that the measures were collected together, not that they moved in lockstep [1]. For a reader, the practical takeaway is that convergence across melatonin, cortisol, and EEG under office fluorescent lamps is an open question rather than a demonstrated pattern.
Light color matters, but the response depends on the person
A separate study using fluorescent lamps at 6500 K or 2500 K, plus incandescent light at the same intensity of about 40 lux in the evening, measured salivary melatonin, subjective sleepiness, and waking EEG [5]. It found that melatonin suppression and the alerting response to blue-enriched light depended on a polymorphism in the clock gene PER3 [5]. In other words, the same light produced different hormonal and EEG outcomes in different people, which is a direct reason the measures may diverge across a study population even when the lamp is identical.
That study did not measure cortisol, and it used evening light rather than an office workday [5]. Its relevance is narrower than the lamp-type question: it shows that genetic variation can decouple melatonin from EEG alertness, so pooling people into one average can hide opposing responses.
Hormone rhythms drift on their own schedule
Studies of shift workers show that melatonin and cortisol rhythms can adapt at different rates to the same schedule. Among police officers working rotating shifts in dim-light environments, 30% adapted to a night-oriented schedule by the melatonin metabolite aMT6s, while 24% adapted by cortisol, and the phase shifts were larger in those who adapted than in those who did not [4]. The two hormones were broadly consistent in direction but not identical in proportion, which is a useful reminder that even two hormone markers of the same system do not move as one.
This evidence comes from field shift-work conditions, not office fluorescent lamps, so it speaks to how hormone measures behave relative to each other rather than to lamp type [4]. It does support the broader point that convergence should be tested rather than assumed. A dog study likewise found that oral melatonin lowered cortisol after surgery [3], but that is a pharmacological effect in animals, not a light effect in office workers, and it does not involve EEG.
Sources used in this answer
Investigating the Effects of Metameric Light on EEG Spectral Measures
A study measured melatonin, cortisol, EEG, ECG, and subjective comfort in healthy humans under two fluorescent lamp types at two intensities, but the available abstract does not report whether these measures converged.
Circadian Biomarkers in Humans: Methodological Insights into the Detection of Melatonin and Cortisol
Melatonin and cortisol are circadian biomarkers whose measurement depends on matrix, assay, ambient light, posture, and sampling time, so standardized protocols are needed for reliable results.
The effects of melatonin on the concentrations of inflammatory cytokines and proteins, serotonin, cortisol and melatonin in ovariohysterectomised female dogs
In ovariohysterectomised female dogs, oral melatonin increased melatonin and serotonin and decreased cortisol in the melatonin-plus-surgery group compared with surgery alone.
Circadian Adaptation of Melatonin and Cortisol in Police Officers Working Rotating Shifts
Among police officers on rotating shifts, about 30% adapted to night schedules by aMT6s and 24% by cortisol, with larger phase delays in adapted than non-adapted officers.
Human Melatonin and Alerting Response to Blue-Enriched Light Depend on a Polymorphism in the Clock Gene PER3
Melatonin suppression and the alerting response to blue-enriched fluorescent light depended on a PER3 polymorphism, with melatonin and waking EEG measured but cortisol not included.